Performance of different UNIFAC parameter sets in describing experimental liquid–liquid equilibrium data of biodiesel systems

Abstract

Several liquid phases with different compositions are generated throughout the biodiesel production chain. The complexity of the fatty compounds involved in the systems can impairs the process simulation and its thermodynamic comprehension. Fortunately, experimental data attained in this study revealed that the fatty classes can be simplified to pseudocomponents in the liquid–liquid equilibrium LLE since their compositions did not change between the liquid phases. Such approach provides a good description of the LLE behavior and benefits the computational simulation of the different steps involved in the biodiesel production and purification. Predictive models as the UNIFAC can used in simulations without the need of exhaustive experimental assays. The predictive capacity of four UNIFAC parameters sets found in the literature was evaluated assuming LLE data of dissimilar fatty systems composed of vegetable oils, partial acylglycerols, free fatty acid, ethyl esters, and ethanol. The UNIFAC-BESSA parameter set exhibited a satisfactory performance in estimating the LLE data with low deviations (1.55 – 4.56%), and can be thus used in computational simulations and thermodynamic modeling of biodiesel systems.